Activity for Victor Mote
-
Victor Mote committed [r13346] on Code
Improvements to natural language dictionaries.
-
Rough-in handling of new "marker" element.
-
Style changes only.
-
1. Move inside class "Element" to a first-class class. 2. Handle new "flow" element in the spell-checking.
-
Move "marker" attributes to an orthogonal element. Add flow element.
-
Improve order of constructor parameters.
-
Add handling for xpath and marker attributes in the intermediate file.
-
Add "marker" attribute for data from the document backing the XML.
-
Store the line and column info in the element stack instead of globally.
-
Improvements to location information for spell-checking documents.
-
Compute the line and column number of each token, relative to the input item container.
-
Add line and column number to the Token interface.
-
Split tracking of location information into its components.
-
Adapt to aXSL change providing more explicit location information.
-
Make location information more explicit, for better downstream use.
-
Allow spaces in most English-Latin patterns, to properly handle open compound words.
-
Improvements to dictionaries and orthographies.
-
Improvements to dictionaries and orthographies.
-
Show location of words in unconfigured orthographies.
-
Improvements to dictionaries and orthographies.
-
Upgrade JUnit to 5.10.2 to fix class loading conflict when running from Eclipse.
-
Dictionary improvements.
-
Separate normal ad-hoc dictionaries from free dictionaries, one managed by the orthography, the other by the orthography server.
-
Handle dictionaries parsed outside of the orthography / dictionary resource scheme.
-
Move dictionaryMap and related methods to superclass, so that the variant class can use.
-
Remove no-longer-needed code.
-
Handle possibility of multiple writing system dictionaries being parsed from a single dictionary file.
-
Improvements to dictionaries and orthographies.
-
Add left single quotation mark as possible intraword punctuation, to handle certain configurations of Scottish names, e.g. M‘Culloch.
-
Improvements to dictionaries/orthographies.
-
Improvements to dictionaries and orthographies.
-
Address new checkstyle warnings.
-
Improvements to dictionaries and orthographies.
-
Improvements to dictionaries and orthographies.
-
Yes, the issue is resolved for me. Thank you. Victor
-
Dictionary improvements.
-
I am seeing similar results committing from both Eclipse and TortoiseSVN. I also changed my password to see if that would fix the problem, with no change. I believe the problems started within the last few days. Username: victormote
-
I am seeing similar results committing from both Eclipse and TortoiseSVN. I also changed my password to see if that would fix the problem, with no change. Username: victormote
-
Handle new axsl-dictionary elements.
-
Improvements to dictionaries and orthographies.
-
Turn word-placeholder attributes into elements.
-
Add word-placeholder element for axsl-dictionary.
-
Move dictionary-related instance variables into the inner class to better separate them from other parser variables.
-
Remove no-longer-needed instance variable.
-
Replace some uses of Logger with Exceptions.
-
Reset last word when starting a new axsl-dictionary element.
-
Remove hard-coded paths to orthography resources by handling relative URLs.
-
Minor doc cleanup.
-
1. Move the main logic for Orthography4a and subclasses to the abstract superclass, leaving the subclasses to basically feed configuration it. 2. Add abstract method that allows the variant class to add the dictionary of the wrapped orthography to the stack that is being processed.
-
Conform to aXSL change: Remove passage of ad-hoc dictionaries when searching for words.
-
Remove passage of ad-hoc dictionaries when searching for words.
-
Allow orthography server to be overridden by a document-specific variant.
-
Rename class in preparation for extraction of an abstract superclass.
-
Move more NaturalLanguage code to the attic (some coming from aXSL).
-
Move more NaturalLanguage code to the attic.
-
Move NaturalLanguage code to the attic.
-
Create class that can override a standard orthography.
-
Minor dependency cleanup.
-
Conform to aXSL changes: Move the text token flow interfaces from axsl-orthography to axsl-fotree.
-
Move the text token flow interfaces from axsl-orthography to axsl-fotree.
-
Conform to aXSL change: Consolidate tokenization concepts into the Lexer.
-
Consolidate tokenization concepts into the Lexer.
-
Conform to aXSL changes: Remove org.axsl.orthography.optional package, moving its interfaces up a level.
-
Remove org.axsl.orthography.optional package, moving its interfaces up a level.
-
Since there is only one lexer now, cache it as an instance variable.
-
Dictionary and orthography improvements.
-
Move creation of actual dictionary to the endElement method, to enable parsing of multiple dictionaries from one file.
-
Dictionary improvements.
-
Minor cleanup.
-
Remove no-longer-needed token type START.
-
Enable test of a token item ending with leading punctuation.
-
Bracket token input with BREAK token types during token resolution, instead of END.
-
Add disabled test for leading punctuation immediately before a new input item.
-
Remove exception for empty content. It should just be ignored. Add exception for null content or writing system for fail-fast.
-
Simplify the locking and processing sequence.
-
Use the token writing system, not the last parsed one.
-
Fix process of getting the orthography.
-
Clean up timing of running the Lexer.
-
Simply ignore empty content. Throw exception for null parameters.
-
Dictionary improvements.
-
Create class to store the breaks in.
-
Enable test that now passes.
-
When resolving ambiguous token types, process all input items as if they were part of a single sequence.
-
Prevent empty content from being added to the Lexer.
-
1. Allow implementations to reject empty content. 2. Improve some variable names.
-
Defer token creation until all breaks have been computed.
-
Minor cleanup.
-
Store the intermediate lexing results with each input item, so that all input items can be processed as contiguous items.
-
Remove some no-longer-needed context parameters. Pass the input index around instead of the input item it references.
-
Create the tokens immediately instead of accumulating their components.
-
Add test of adjacent tokens in different writing systems.
-
Dictionary improvements.
-
Conform to aXSL change: Rename token type WHITESPACE to BREAK, and add documentation about why.
-
Rename token type WHITESPACE to BREAK, and add documentation about why.
-
Handle an embedded slash or solidus.
-
Resolve ambiguous trailing punctuation that can be resolved from its context.
-
Handle transient trailing punctuation that is followed by resolved trailing punctuation.
-
Handle lexing of horizontal ellipsis.
-
If a word is not found in the dictionary, but is following by ambiguous trailing punctuation, append the punctuation and try again.
-
Conform to aXSL change: Remove explicit-token concept. This is now considered to be out-of-scope for a lexer, but should be handled in a dictionary.
1 >